Aims: Reproductive problems (such as infertility, miscarriage, and birth defects) have been linked to environmental toxins in laboratory animals. However, little is known about similar effects in humans. Human reproduction is not only a likely target for environmental toxins, but it may be a more sensitive endpoint than cancer or other chronic disease, which can require many years to become apparent. Thus, reproductive problems may serve as an early signal of toxic exposures. Our goal is to describe the basic biology of human reproduction, to develop improved epidemiologic tools for detecting environmental damage to reproduction, and to identify specific environmental factors that interfere with human reproduction. Procedures and techniques: Urinary hCG and steroid assays, new statistical methods, vital statistics, special cohort studies, and national registries, Scandinavian birth registries, maternally-linked pregnancy records in Norway, repeated observations of risk within one person Accomplishments: We combined our data on the fertile window (comprising the five days before ovulation and the day of ovulation) with the natural variability in the timing of ovulation across menstrual cycles to estimate the distribution of fertile days in the menstrual cycle. The fertile days were markedly dispersed, including a risk of conception as late as the fifth or sixth weeks of the menstrual cycle. We used the daily diary reports in our study to describe onset of symptoms among women with subsequently successful pregnancies, and among women whose pregnancies ended in loss. Even the early losses (within 6 weeks of the last menstrual period) were found to produce occasional symptoms of pregnancy, demonstrating that early losses are not all literally "sub-clinical". Using data from the Norwegian Medical Birth Registry, we have found that women with new partners have, on average, a longer interval between the two pregnancies. To our surprise, we also found that risk of preeclampsia rises with longer interval. The risk with the new partner actually is decreased after adjusting for pregnancy interval. This increasing risk with pregnancy interval suggests a previously unrecognized causal mechanism for preeclampsia, and raises doubts about the immune theory of preeclampsia etiology.